To see the other types of publications on this topic, follow the link: Application to image restoration.
Dissertations / Theses on the topic 'Application to image restoration'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Application to image restoration.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Boukouvala, Erisso. "Image restoration techniques and application on astronomical images." Thesis, University of Reading, 2004. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.414571.
Full text
2
Qiu, Zhen. "Feature-preserving image restoration and its application in biological fluorescence microscopy." Thesis, Heriot-Watt University, 2013. http://hdl.handle.net/10399/2682.
Full textAbstract:
This thesis presents a new investigation of image restoration and its application to fluorescence cell microscopy. The first part of the work is to develop advanced image denoising algorithms to restore images from noisy observations by using a novel featurepreserving diffusion approach. I have applied these algorithms to different types of images, including biometric, biological and natural images, and demonstrated their superior performance for noise removal and feature preservation, compared to several state of the art methods. In the second part of my work, I explore a novel, simple and inexpensive super-resolution restoration method for quantitative microscopy in cell biology. In this method, a super-resolution image is restored, through an inverse process, by using multiple diffraction-limited (low) resolution observations, which are acquired from conventional microscopes whilst translating the sample parallel to the image plane, so referred to as translation microscopy (TRAM). A key to this new development is the integration of a robust feature detector, developed in the first part, to the inverse process to restore high resolution images well above the diffraction limit in the presence of strong noise. TRAM is a post-image acquisition computational method and can be implemented with any microscope. Experiments show a nearly 7-fold increase in lateral spatial resolution in noisy biological environments, delivering multi-colour image resolution of ~30 nm.
3
Abboud, Feriel. "Restoration super-resolution of image sequences : application to TV archive documents." Thesis, Paris Est, 2017. http://www.theses.fr/2017PESC1038/document.
Full textAbstract:
Au cours du dernier siècle, le volume de vidéos stockées chez des organismes tel que l'Institut National de l'Audiovisuel a connu un grand accroissement. Ces organismes ont pour mission de préserver et de promouvoir ces contenus, car, au-delà de leur importance culturelle, ces derniers ont une vraie valeur commerciale grâce à leur exploitation par divers médias. Cependant, la qualité visuelle des vidéos est souvent moindre comparée à celles acquises par les récents modèles de caméras. Ainsi, le but de cette thèse est de développer de nouvelles méthodes de restauration de séquences vidéo provenant des archives de télévision française, grâce à de récentes techniques d'optimisation. La plupart des problèmes de restauration peuvent être résolus en les formulant comme des problèmes d'optimisation, qui font intervenir plusieurs fonctions convexes mais non-nécessairement différentiables. Pour ce type de problèmes, on a souvent recourt à un outil efficace appelé opérateur proximal. Le calcul de l'opérateur proximal d'une fonction se fait de façon explicite quand cette dernière est simple. Par contre, quand elle est plus complexe ou fait intervenir des opérateurs linéaires, le calcul de l'opérateur proximal devient plus compliqué et se fait généralement à l'aide d'algorithmes itératifs. Une première contribution de cette thèse consiste à calculer l'opérateur proximal d'une somme de plusieurs fonctions convexes composées avec des opérateurs linéaires. Nous proposons un nouvel algorithme d'optimisation de type primal-dual, que nous avons nommé Algorithme Explicite-Implicite Dual par Blocs. L'algorithme proposé permet de ne mettre à jour qu'un sous-ensemble de blocs choisi selon une règle déterministe acyclique. Des résultats de convergence ont été établis pour les deux suites primales et duales de notre algorithme. Nous avons appliqué notre algorithme au problème de déconvolution et désentrelacement de séquences vidéo. Pour cela, nous avons modélisé notre problème sous la forme d'un problème d'optimisation dont la solution est obtenue à l'aide de l'algorithme explicite-implicite dual par blocs. Dans la deuxième partie de cette thèse, nous nous sommes intéressés au développement d'une version asynchrone de notre l'algorithme explicite-implicite dual par blocs. Dans cette nouvelle extension, chaque fonction est considérée comme locale et rattachée à une unité de calcul. Ces unités de calcul traitent les fonctions de façon indépendante les unes des autres. Afin d'obtenir une solution de consensus, il est nécessaire d'établir une stratégie de communication efficace. Un point crucial dans le développement d'un tel algorithme est le choix de la fréquence et du volume de données à échanger entre les unités de calcul, dans le but de préserver de bonnes performances d'accélération. Nous avons évalué numériquement notre algorithme distribué sur un problème de débruitage de séquences vidéo. Les images composant la vidéo sont partitionnées de façon équitable, puis chaque processeur exécute une instance de l'algorithme de façon asynchrone et communique avec les processeurs voisins. Finalement, nous nous sommes intéressés au problème de déconvolution aveugle, qui vise à estimer le noyau de convolution et la séquence originale à partir de la séquence dégradée observée. Nous avons proposé une nouvelle méthode basée sur la formulation d'un problème non-convexe, résolu par un algorithme itératif qui alterne entre l'estimation de la séquence originale et l'identification du noyau. Notre méthode a la particularité de pouvoir intégrer divers types de fonctions de régularisations avec des propriétés mathématiques différentes. Nous avons réalisé des simulations sur des séquences synthétiques et réelles, avec différents noyaux de convolution. La flexibilité de notre approche nous a permis de réaliser des comparaisons entre plusieurs fonctions de régularisation convexes et non-convexes, en terme de qualité d'estimationThe last century has witnessed an explosion in the amount of video data stored with holders such as the National Audiovisual Institute whose mission is to preserve and promote the content of French broadcast programs. The cultural impact of these records, their value is increased due to commercial reexploitation through recent visual media. However, the perceived quality of the old data fails to satisfy the current public demand. The purpose of this thesis is to propose new methods for restoring video sequences supplied from television archive documents, using modern optimization techniques with proven convergence properties. In a large number of restoration issues, the underlying optimization problem is made up with several functions which might be convex and non-necessarily smooth. In such instance, the proximity operator, a fundamental concept in convex analysis, appears as the most appropriate tool. These functions may also involve arbitrary linear operators that need to be inverted in a number of optimization algorithms. In this spirit, we developed a new primal-dual algorithm for computing non-explicit proximity operators based on forward-backward iterations. The proposed algorithm is accelerated thanks to the introduction of a preconditioning strategy and a block-coordinate approach in which at each iteration, only a "block" of data is selected and processed according to a quasi-cyclic rule. This approach is well suited to large-scale problems since it reduces the memory requirements and accelerates the convergence speed, as illustrated by some experiments in deconvolution and deinterlacing of video sequences. Afterwards, a close attention is paid to the study of distributed algorithms on both theoretical and practical viewpoints. We proposed an asynchronous extension of the dual forward-backward algorithm, that can be efficiently implemented on a multi-cores architecture. In our distributed scheme, the primal and dual variables are considered as private and spread over multiple computing units, that operate independently one from another. Nevertheless, communication between these units following a predefined strategy is required in order to ensure the convergence toward a consensus solution. We also address in this thesis the problem of blind video deconvolution that consists in inferring from an input degraded video sequence, both the blur filter and a sharp video sequence. Hence, a solution can be reached by resorting to nonconvex optimization methods that estimate alternatively the unknown video and the unknown kernel. In this context, we proposed a new blind deconvolution method that allows us to implement numerous convex and nonconvex regularization strategies, which are widely employed in signal and image processing
4
Al-Suwailem, Umar A. "Continuous spatial domain image identification and restoration with multichannel applications /." free to MU campus, to others for purchase, 1996. http://wwwlib.umi.com/cr/mo/fullcit?p9737865.
Full text
5
Auyeung, Cheung. "Optimal constraint-based signal restoration and its applications." Diss., Georgia Institute of Technology, 1988. http://hdl.handle.net/1853/15785.
Full text
6
Eastlick, Anne C. "Genre criticism : an application of BP's image restoration campaign to the crisis communication genre." Scholarly Commons, 2011. https://scholarlycommons.pacific.edu/uop_etds/767.
Full textAbstract:
Within two months of its emergence, the BP Gulf Oil spill had become the worst environmental disaster in United States history. However, for those studying public relations the oil spill brought more than ecological disaster, by providing a case study of crisis communication. Although there were a number of crisis responses from BP throughout the course of the oil spill, the primary crisis response crafted by BP was an image restoration campaign which premiered in early June 2010. This campaign, though it exhibits qualities of a standard crisis response, was wildly unpopular with the United States Government and citizenry. This rhetorical analysis attempts to uncover the reasons behind the campaign's failure through an application of the genre model of criticism. By defining the crisis communication genre and applying it to the artifact, the current study uncovers the reasons behind the failure of the campaign. Through this discussion, this analysis identifies that BP did not address all necessary exigencies, nor did it consider the influence a rhetor can have on a message. An explanation for the failure of BP' s campaign provided a plethora of implications to the fields of public . relations and rhetorical criticism, while beginning a discussion to help define the crisis communication genre.
7
Wen, Youwei. "Fast solvers for Toeplitz systems with applications to image restoration." Click to view the E-thesis via HKUTO, 2006. http://sunzi.lib.hku.hk/hkuto/record/B3688280X.
Full text
8
Wen, Youwei, and 文有為. "Fast solvers for Toeplitz systems with applications to image restoration." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2006. http://hub.hku.hk/bib/B3688280X.
Full text
9
Saeed, Mohammed. "Maximum likelihood parameter estimation of mixture models and its application to image segmentation and restoration." Thesis, Massachusetts Institute of Technology, 1997. http://hdl.handle.net/1721.1/43410.
Full text
10
Gibbs, Alison L. "Convergence of Markov chain Monte Carlo algorithms with applications to image restoration." Thesis, National Library of Canada = Bibliothèque nationale du Canada, 2000. http://www.collectionscanada.ca/obj/s4/f2/dsk2/ftp03/NQ50003.pdf.
Full text
11
Ahmadvand, Samaneh. "Efficient Visibility Restoration Method Using a Single Foggy Image in Vehicular Applications." Thesis, Université d'Ottawa / University of Ottawa, 2018. http://hdl.handle.net/10393/38486.
Full textAbstract:
Foggy and hazy weather conditions considerably effect visibility distance which impacts speed, flow of traffic, travel time delay and increases the risk accidents. Bad weather condition is considered a cause of road accidents, since it the poor conditions can effect drivers field of vision. In addition, fog, haze and mist can have negative influences on visual applications in the open air since they decrease visibility by lowering the contrast and whitening the visible color palette. The poor visibility in these images leads to some failures in recognition and detection of the outdoor object systems and also in Intelligent Transportation Systems (ITS). In this thesis, we propose an image visibility restoration algorithm under foggy weather in intelligent transportation systems. Various camera based Advanced Driver Assistant Systems (ADAS), which can be improved by applying the visibility restoration algorithm, have been applied in this field of study to enhance vehicle safety by displaying the image from a frontal camera to driver after visibility enhancement.
To remove fog, automatic methods have been proposed which are categorized into two approaches based on the number of input images: 1) methods which are using polarizing filters, 2) methods which are using captured images from different fog densities. In both of these approaches multiple images are required which have to be taken from exactly the same point of view. While these applications can generate good results, their requirements make them impractical, particularly in real time applications, such as intelligent transportation systems. Therefore, in this thesis we introduce a high-performance visibility restoration algorithm only using a single foggy image which applies a recursive filtering to preserve the edge of images and videos in real time and also compute depth map of the scene to restore image. The applied edge preserving filtering is based on a domain transform in which 1-Dimensional edge-preserving filtering is performed by preserving the geodesic distance between points on the curves that is adaptable with wrapping the input signal. The proposed algorithm can be applied in intelligent transportation system applications, such as Advanced Driver Assistance Systems (ADAS). The main features of the proposed algorithm are its speed, which plays a main role in real time applications, since 1-Dimensional operations are used in the applied filtering leads to remarkable speedups in comparison with classical median filters and robust bilateral lfilters. Potential of memory saving is considered as another one advantage of the proposed model and also the parameters of applied edge-preserving filtering do not effect on its computational cost. It is the first edge-preserving filter for color images with arbitrary scales in real time. The proposed algorithm is also able to handle both color and gray-level images and achieves the restored image without the presence of artifacts in comparison with other state-of-the-art algorithms.
12
Kou, Kit Ian. "Fast transform based operators for Toeplitz systems and their applications in image restoration." Thesis, University of Macau, 1999. http://umaclib3.umac.mo/record=b1446619.
Full text
13
Esser, John Ernest. "Primal dual algorithms for convex models and applications to image restoration, registration and nonlocal inpainting." Diss., Restricted to subscribing institutions, 2010. http://proquest.umi.com/pqdweb?did=2023768061&sid=1&Fmt=2&clientId=1564&RQT=309&VName=PQD.
Full text
14
Brito, Loeza Carlos Francisco. "Fast numerical algorithms for high order partial differential equations with applications to image restoration techniques." Thesis, University of Liverpool, 2009. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.526786.
Full text
15
Pang, Ho-Yuen. "Novel super-resolution algorithms and enhanced noise removal algorithm for image restoration systems and applications." Diss., The University of Arizona, 2002. http://hdl.handle.net/10150/279978.
Full textAbstract:
This dissertation is concerned with the introduction of a systematic way of modeling image processing. A dynamic imaging system model constructed from an information theory framework is proposed. Unlike an earlier simple model, the proposed dynamic imaging system (DIS) model is suitable for a wide range of applications. This DIS model is inspired by the Shannon communication theory. The Shannon communication theory is credited for the rapid development of the communication industry. Currently, most image processing researchers focus on developing fast algorithms and better hardware. An information theoretic-based approach to image processing could bring as large an impact to the image processing area as Shannon's communication theory had on the communications area. This proposed DIS model will use the information obtained from the acquired images to provide an estimation of the unknown atmospheric turbulence, vibration, etc. It will also automatically adjust the sampling rate, wavelength band, and algorithms of choice, to produce the best possible restored image with limited information under uncertainty. This dissertation develops the concept of the DIS model including its basic components. We have implemented three parts of this system. First, we implemented a noise removal algorithm based on the Markov random field (MRF). It is shown that this algorithm achieves better performance than other MRF-based algorithms in noise removal. Second, we have implemented a hybrid maximum likelihood/projection-on-convex-set image restoration algorithm and demonstrate that it outperforms the maximum likelihood algorithm. Third, we have implemented a self-organized map-based image restoration algorithm and compare its performance to several well-known methods. It can be implemented in parallel processing to achieve super-resolution in real time without performing a time consuming iteration process. The impact of the development of these DIS system critical components is discussed and future research areas are elucidated.
16
Lauga, Guillaume. "Méthodes proximales multi-niveaux et application à la restauration d'images." Electronic Thesis or Diss., Lyon, École normale supérieure, 2024. http://www.theses.fr/2024ENSL0089.
Full textAbstract:
La taille des problèmes de restauration d'images ne fait qu'augmenter. Cette croissance pose un problème majeur de passage à l'échelle pour les algorithmes d'optimisation, qui peinent à fournir des solutions satisfaisantes en un temps raisonnable. Parmi les méthodes proposées pour surmonter ce défi, les méthodes multi-niveaux semblent être un candidat idéal. En réduisant de manière systématique la dimension du problème, le coût computationnel nécessaire à sa résolution peut diminuer drastiquement. Ce type d'approche est classique pour la résolution numérique des équations aux dérivées partielles (EDP), avec des garanties théoriques et des démonstrations pratiques pour expliquer leur succès. Cependant, les méthodes actuelles d'optimisation multi-niveaux n'ont pas les mêmes garanties, ni les mêmes performances. Dans cette thèse, nous proposons de combler une partie de cet écart en introduisant un nouvel algorithme multi-niveaux, IML FISTA, possédant les garanties de convergence théoriques optimales pour les problèmes d'optimisation convexes non-lisses, i.e., convergence vers un minimiseur et taux de convergence de la fonction objectif vers une valeur minimale. IML FISTA est aussi en mesure de traiter les régularisations de l'état-de-l'art en restauration d'images. En comparant IML FISTA aux algorithmes standard sur un grand nombre de problèmes de restauration d'images: défloutage, débruitage, reconstruction de pixels manquants pour des images en couleur et des images hyperspectrales, ainsi qu'en reconstruction d'images radio-interférométriques, nous montrons qu'IML FISTA est capable d'accélérer la résolution de ces problèmes de manière significative. Le cadre d'IML FISTA est suffisamment général pour s'adapter à de nombreux autres problèmes de restauration d'images. Nous concluons cette thèse en proposant un nouveau point de vue sur les algorithmes multi-niveaux, en démontrant leur équivalence, dans certains cas, avec les algorithmes de descente par coordonnées qui sont nettement plus étudiés dans la littérature de l'optimisation non-lisse. Ce nouveau cadre théorique nous permet d'analyser les algorithmes multi-niveaux de manière plus rigoureuse, et notamment d'étendre leurs garanties de convergence à des problèmes non-lisses et non-convexes. Ce cadre est moins général que celui d'IML FISTA, mais il ouvre la voie à une conception plus solide sur le plan théorique des algorithmes multi-niveauxThe size of image restoration problems is constantly increasing. This growth poses a major scaling problem for optimisation algorithms, which struggle to provide satisfactory solutions in a reasonable amount of time. Among the methods proposed to overcome this challenge, multilevel methods seem to be an ideal candidate. By systematically reducing the size of the problem, the computational cost of solving it can be drastically reduced. This type of approach is standard in the numerical solution of partial differential equations (PDEs), with theoretical guarantees and practical demonstrations to explain their success. However, current multilevel optimisation methods do not have the same guarantees nor the same performance. In this thesis, we propose to bridge part of this gap by introducing a new multilevel algorithm, IML FISTA, which has the optimal theoretical convergence guarantees for convex non-smooth optimisation problems, i.e. convergence to a minimiser and convergence rate of the objective function to a minimum value. IML FISTA is also able to handle state-of-the-art regularisations in image restoration. By comparing IML FISTA with standard algorithms on many image restoration problems: deblurring, denoising, reconstruction of missing pixels for colour and hyperspectral images, and reconstruction of radio-interferometric images, we show that IML FISTA is capable of significantly speeding up the resolution of these problems. As IML FISTA's framework is sufficiently general, it can be adapted to many other image restoration problems. We conclude this thesis by proposing a new point of view on multilevel algorithms, by demonstrating their equivalence, in certain cases, with coordinate descent algorithms, which are much more widely studied in the non-smooth optimisation literature. This new theoretical framework allows us to analyse multi-level algorithms more rigorously, and in particular to extend their convergence guarantees to non-smooth and non-convex problems. This framework is less general than that of IML FISTA, but it paves the way for a more theoretically robust design of multilevel algorithms
17
Heinrich, André. "Fenchel duality-based algorithms for convex optimization problems with applications in machine learning and image restoration." Doctoral thesis, Universitätsbibliothek Chemnitz, 2013. http://nbn-resolving.de/urn:nbn:de:bsz:ch1-qucosa-108923.
Full textAbstract:
The main contribution of this thesis is the concept of Fenchel duality with a focus on its application in the field of machine learning problems and image restoration tasks. We formulate a general optimization problem for modeling support vector machine tasks and assign a Fenchel dual problem to it, prove weak and strong duality statements as well as necessary and sufficient optimality conditions for that primal-dual pair. In addition, several special instances of the general optimization problem are derived for different choices of loss functions for both the regression and the classifification task. The convenience of these approaches is demonstrated by numerically solving several problems. We formulate a general nonsmooth optimization problem and assign a Fenchel dual problem to it. It is shown that the optimal objective values of the primal and the dual one coincide and that the primal problem has an optimal solution under certain assumptions. The dual problem turns out to be nonsmooth in general and therefore a regularization is performed twice to obtain an approximate dual problem that can be solved efficiently via a fast gradient algorithm. We show how an approximate optimal and feasible primal solution can be constructed by means of some sequences of proximal points closely related to the dual iterates. Furthermore, we show that the solution will indeed converge to the optimal solution of the primal for arbitrarily small accuracy. Finally, the support vector regression task is obtained to arise as a particular case of the general optimization problem and the theory is specialized to this problem. We calculate several proximal points occurring when using difffferent loss functions as well as for some regularization problems applied in image restoration tasks. Numerical experiments illustrate the applicability of our approach for these types of problems.
18
Karch, Barry K. "Improved Super-Resolution Methods for Division-of-Focal-Plane Systems in Complex and Constrained Imaging Applications." University of Dayton / OhioLINK, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=dayton1429032650.
Full text
19
Wang, Chong, and 王翀. "Joint color-depth restoration with kinect depth camera and its applications to image-based rendering and hand gesture recognition." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2014. http://hdl.handle.net/10722/206343.
Full text
20
Couprie, Camille. "Graph-based variational optimization and applications in computer vision." Phd thesis, Université Paris-Est, 2011. http://tel.archives-ouvertes.fr/tel-00666878.
Full textAbstract:
Many computer vision applications such as image filtering, segmentation and stereovision can be formulated as optimization problems. Recently discrete, convex, globally optimal methods have received a lot of attention. Many graph-based methods suffer from metrication artefacts, segmented contours are blocky in areas where contour information is lacking. In the first part of this work, we develop a discrete yet isotropic energy minimization formulation for the continuous maximum flow problem that prevents metrication errors. This new convex formulation leads us to a provably globally optimal solution. The employed interior point method can optimize the problem faster than the existing continuous methods. The energy formulation is then adapted and extended to multi-label problems, and shows improvements over existing methods. Fast parallel proximal optimization tools have been tested and adapted for the optimization of this problem. In the second part of this work, we introduce a framework that generalizes several state-of-the-art graph-based segmentation algorithms, namely graph cuts, random walker, shortest paths, and watershed. This generalization allowed us to exhibit a new case, for which we developed a globally optimal optimization method, named "Power watershed''. Our proposed power watershed algorithm computes a unique global solution to multi labeling problems, and is very fast. We further generalize and extend the framework to applications beyond image segmentation, for example image filtering optimizing an L0 norm energy, stereovision and fast and smooth surface reconstruction from a noisy cloud of 3D points
21
Wang, Shanshan. "Study of analytic and trained dictionaries for sparse representation and its applications." Thesis, The University of Sydney, 2014. http://hdl.handle.net/2123/11486.
Full textAbstract:
Inverse problems are ubiquitous in the field of medical imaging and image processing. Prominent examples include image reconstruction and image denoising. The goal of these problems is to reconstruct or restore an unknown image from a set of direct/indirect measurements. However, due to the limitation of the acquisition time or the existence of noise, the obtained measurements are often corrupted or incomplete, which introduces big challenges for the reconstruction process. In order to remove the noise or overcome the ill-posed nature caused by the insufficient measurements, it is necessary to explore the prior knowledge and utilize this to form constraints in the reconstruction process so as to make up for the missing or corrupted information. However, traditional prior knowledge regularizations and their corresponding algorithms suffer from loss of the detailed information such as texture and structure while reducing the image degradation factors. To this end, based on the physiological findings about human visual system (HVS), this thesis focuses on exploiting the prior knowledge of self-similarity and sparsity inherited in the image and has developed a series of novel algorithms via sparse representation over analytic and trained dictionaries. The main work and contributions are summarized as follows: 1) A Gabor feature based nonlocal means (GFNLM) algorithm for textured image restoration; 2) an adaptive dictionary learning based impulse noise removal (DL-INR) algorithm; 3) a Fenchel duality based dictionary learning (FD-DL) algorithm and 4) a spatially adaptive constrained dictionary learning (SAC-DL) algorithm for Rician noise removal and a joint entropy regularized bias removal approach.
22
Heinrich, André [Verfasser], Gert [Akademischer Betreuer] Wanka, Gert [Gutachter] Wanka, and Jörg [Gutachter] Fliege. "Fenchel duality-based algorithms for convex optimization problems with applications in machine learning and image restoration / André Heinrich ; Gutachter: Gert Wanka, Jörg Fliege ; Betreuer: Gert Wanka." Chemnitz : Universitätsbibliothek Chemnitz, 2013. http://d-nb.info/1214245315/34.
Full text
23
Bringer, Yves. "Performances de nouvelles architectures machines pour la mise en oeuvre d'algorithmes de traitement et d'analyse d'image." Saint-Etienne, 1993. http://www.theses.fr/1993STET4024.
Full textAbstract:
Une carte électronique a été réalisée à l'Institut de chimie et physique industrielles de Lyon utilisant quatre processeurs à architecture à flot de données et programmable liant ainsi puissance et souplesse d'utilisation. Pour valider cette architecture pour le traitement et l'analyse d'image, l'approche a été double : - mise en oeuvre d'algorithme à la fois coûteux et originaux scientifiquement : algorithme de Danielson, suppression de flou, reconstruction 3D. - implantation sur site industriel avec prise en compte des contraintes de temps et intégration dans une chaine complète de contrôle
24
Gilardet, Mathieu. "Étude d’algorithmes de restauration d’images sismiques par optimisation de forme non linéaire et application à la reconstruction sédimentaire." Thesis, Pau, 2013. http://www.theses.fr/2013PAUU3040/document.
Full textAbstract:
Nous présentons une nouvelle méthode pour la restauration d'images sismiques. Quand on l'observe, une image sismique est le résultat d'un système de dépôt initial qui a été transformé par un ensemble de déformations géologiques successives (flexions, glissement de la faille, etc) qui se sont produites sur une grande période de temps. L'objectif de la restauration sismique consiste à inverser les déformations pour fournir une image résultante qui représente le système de dépôt géologique tel qu'il était dans un état antérieur. Classiquement, ce procédé permet de tester la cohérence des hypothèses d'interprétations formulées par les géophysiciens sur les images initiales. Dans notre contribution, nous fournissons un outil qui permet de générer rapidement des images restaurées et qui aide donc les géophysiciens à reconnaître et identifier les caractéristiques géologiques qui peuvent être très fortement modifiées et donc difficilement identifiables dans l'image observée d'origine. Cette application permet alors d'assister ces géophysiciens pour la formulation d'hypothèses d'interprétation des images sismiques. L'approche que nous introduisons est basée sur un processus de minimisation qui exprime les déformations géologiques en termes de contraintes géométriques. Nous utilisons une approche itérative de Gauss-Newton qui converge rapidement pour résoudre le système. Dans une deuxième partie de notre travail nous montrons différents résultats obtenus dans des cas concrets afin d'illustrer le processus de restauration d'image sismique sur des données réelles et de montrer comment la version restaurée peut être utilisée dans un cadre d'interprétation géologiqueWe present a new method for seismic image restoration. When observed, a seismic image is the result of an initial deposit system that has been transformed by a set of successive geological deformations (folding, fault slip, etc) that occurred over a large period of time. The goal of seismic restoration consists in inverting the deformations to provide a resulting image that depicts the geological deposit system as it was in a previous state. With our contribution, providing a tool that quickly generates restored images helps the geophysicists to recognize geological features that may be too strongly altered in the observed image. The proposed approach is based on a minimization process that expresses geological deformations in terms of geometrical constraints. We use a quickly-converging Gauss-Newton approach to solve the system. We provide results to illustrate the seismic image restoration process on real data and present how the restored version can be used in a geological interpretation framework
25
Hadj-Youcef, Mohamed Elamine. "Spatio spectral reconstruction from low resolution multispectral data : application to the Mid-Infrared instrument of the James Webb Space Telescope." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS326/document.
Full textAbstract:
Cette thèse traite un problème inverse en astronomie. L’objectif est de reconstruire un objet 2D+λ, ayant une distribution spatiale et spectrale, à partir d’un ensemble de données multispectrales de basse résolution fournies par l’imageur MIRI (Mid-InfraRed Instrument), qui est à bord du prochain télescope spatial James Webb Space Telescope (JWST). Les données multispectrales observées souffrent d’un flou spatial qui dépend de la longueur d’onde. Cet effet est dû à la convolution par la réponse optique (PSF). De plus, les données multi-spectrales souffrent également d’une sévère dégradation spectrale en raison du filtrage spectral et de l’intégration par le détecteur sur de larges bandes. La reconstruction de l’objet original est un problème mal posé en raison du manque important d’informations spectrales dans l’ensemble de données multispectrales. La difficulté se pose alors dans le choix d’une représentation de l’objet permettant la reconstruction de l’information spectrale. Un modèle classique utilisé jusqu’à présent considère une PSF invariante spectralement par bande, ce qui néglige la variation spectrale de la PSF. Cependant, ce modèle simpliste convient que dans le cas d’instrument à une bande spectrale très étroite, ce qui n’est pas le cas pour l’imageur de MIRI. Notre approche consiste à développer une méthode pour l’inversion qui se résume en quatre étapes : (1) concevoir un modèle de l’instrument reproduisant les données multispectrales observées, (2) proposer un modèle adapté pour représenter l’objet à reconstruire, (3) exploiter conjointement l’ensemble des données multispectrales, et enfin (4) développer une méthode de reconstruction basée sur la régularisation en introduisant des priori à la solution. Les résultats de reconstruction d’objets spatio-spectral à partir de neuf images multispectrales simulées de l’imageur de MIRI montrent une augmentation significative des résolutions spatiale et spectrale de l’objet par rapport à des méthodes conventionnelles. L’objet reconstruit montre l’effet de débruitage et de déconvolution des données multispectrales. Nous avons obtenu une erreur relative n’excédant pas 5% à 30 dB et un temps d’exécution de 1 seconde pour l’algorithme de norm-l₂ et 20 secondes avec 50 itérations pour l’algorithme norm-l₂/l₁. C’est 10 fois plus rapide que la solution itérative calculée par l’algorithme de gradient conjuguéThis thesis deals with an inverse problem in astronomy. The objective is to reconstruct a spatio-spectral object, having spatial and spectral distributions, from a set of low-resolution multispectral data taken by the imager MIRI (Mid-InfraRed Instrument), which is on board the next space telescope James Webb Space Telescope (JWST). The observed multispectral data suffers from a spatial blur that varies according to the wavelength due to the spatial convolution with a shift-variant optical response (PSF). In addition the multispectral data also suffers from severe spectral degradations because of the spectral filtering and the integration by the detector over broad bands. The reconstruction of the original object is an ill-posed problem because of the severe lack of spectral information in the multispectral dataset. The difficulty then arises in choosing a representation of the object that allows the reconstruction of this spectral information. A common model used so far considers a spectral shift-invariant PSF per band, which neglects the spectral variation of the PSF. This simplistic model is only suitable for instruments with a narrow spectral band, which is not the case for the imager of MIRI. Our approach consists of developing an inverse problem framework that is summarized in four steps: (1) designing an instrument model that reproduces the observed multispectral data, (2) proposing an adapted model to represent the sought object, (3) exploiting all multispectral dataset jointly, and finally (4) developing a reconstruction method based on regularization methods by enforcing prior information to the solution. The overall reconstruction results obtained on simulated data of the JWST/MIRI imager show a significant increase of spatial and spectral resolutions of the reconstructed object compared to conventional methods. The reconstructed object shows a clear denoising and deconvolution of the multispectral data. We obtained a relative error below 5% at 30 dB, and an execution time of 1 second for the l₂-norm algorithm and 20 seconds (with 50 iterations) for the l₂/l₁-norm algorithm. This is 10 times faster than the iterative solution computed by conjugate gradients
26
Ungan, Cahit Ugur. "Nonlinear Image Restoration." Master's thesis, METU, 2005. http://etd.lib.metu.edu.tr/upload/2/12606796/index.pdf.
Full textAbstract:
This thesis analyzes the process of deblurring of degraded images generated by space-variant nonlinear image systems with Gaussian observation noise. The restoration of blurred images is performed by using two methodsa modified version of the Optimum Decoding Based Smoothing Algorithm and the Bootstrap Filter Algorithm which is a version of Particle Filtering methods. A computer software called MATLAB is used for performing the simulations of image estimation. The results of some simulations for various observation and image models are presented.
27
Dolne, Jean J. "Estimation theoretical image restoration." Thesis, Massachusetts Institute of Technology, 2008. http://hdl.handle.net/1721.1/47859.
Full textAbstract:
Thesis (S.M.)--Massachusetts Institute of Technology, System Design and Management Program, 2008.Includes bibliographical references.
In this thesis, we have developed an extensive study to evaluate image restoration from a single image, colored or monochromatic. Using a mixture of Gaussian and Poisson noise process, we derived an objective function to estimate the unknown object and point spread function (psf) parameters. We have found that, without constraint enforcement, this blind deconvolution algorithm tended to converge to the trivial solution: delta function as the estimated psf and the detected image as the estimated object. We were able to avoid this solution set by enforcing a priori knowledge about the characteristics of the solution, which included the constraints on object sharpness, energy conservation, impulse response point spread function solution, and object gradient statistics. Applying theses constraints resulted in significantly improved solutions, as evaluated visually and quantitatively using the distance of the estimated to the true function. We have found that the distance of the estimated psf was correlated better with visual observation than the distance metric using the estimated object. Further research needs to be done in this area. To better pose the problem, we expressed the point spread function as a series of Gaussian basis functions, instead of the pixel basis function formalism used above. This procedure has reduced the dimensionality of the parameter space and has resulted in improved results, as expected. We determined a set of weights that yielded optimum algorithm performance.
(cont.) Additional research needs to be done to include the weight set as optimization parameters. This will free the user from having to adjust the weights manually. Of course, if certain knowledge of a weight is available, then it may be better to start with that as an initial guess and optimize from there. With the knowledge that the gradient of the object obeys long-tailed distribution, we have incorporated a constraint using the first two moments, mean and variance, of the gradient of the object in the objective function. Additional research should be done to incorporate the entire distribution in the objective and gradient functions and evaluate the performance.
by Jean J. Dolne.
S.M.
28
Pai, Hung-ta. "Multichannel blind image restoration /." Digital version accessible at:, 1999. http://wwwlib.umi.com/cr/utexas/main.
Full text
29
Reichenbach, Stephen Edward. "Small-kernel image restoration." W&M ScholarWorks, 1989. https://scholarworks.wm.edu/etd/1539623783.
Full textAbstract:
The goal of image restoration is to remove degradations that are introduced during image acquisition and display. Although image restoration is a difficult task that requires considerable computation, in many applications the processing must be performed significantly faster than is possible with traditional algorithms implemented on conventional serial architectures. as demonstrated in this dissertation, digital image restoration can be efficiently implemented by convolving an image with a small kernel. Small-kernel convolution is a local operation that requires relatively little processing and can be easily implemented in parallel. A small-kernel technique must compromise effectiveness for efficiency, but if the kernel values are well-chosen, small-kernel restoration can be very effective.;This dissertation develops a small-kernel image restoration algorithm that minimizes expected mean-square restoration error. The derivation of the mean-square-optimal small kernel parallels that of the Wiener filter, but accounts for explicit spatial constraints on the kernel. This development is thorough and rigorous, but conceptually straightforward: the mean-square-optimal kernel is conditioned only on a comprehensive end-to-end model of the imaging process and spatial constraints on the kernel. The end-to-end digital imaging system model accounts for the scene, acquisition blur, sampling, noise, and display reconstruction. The determination of kernel values is directly conditioned on the specific size and shape of the kernel. Experiments presented in this dissertation demonstrate that small-kernel image restoration requires significantly less computation than a state-of-the-art implementation of the Wiener filter yet the optimal small-kernel yields comparable restored images.;The mean-square-optimal small-kernel algorithm and most other image restoration algorithms require a characterization of the image acquisition device (i.e., an estimate of the device's point spread function or optical transfer function). This dissertation describes an original method for accurately determining this characterization. The method extends the traditional knife-edge technique to explicitly deal with fundamental sampled system considerations of aliasing and sample/scene phase. Results for both simulated and real imaging systems demonstrate the accuracy of the method.
30
Katsaggelos, Aggelos Konstantinos. "Constrained iterative image restoration algorithms." Diss., Georgia Institute of Technology, 1985. http://hdl.handle.net/1853/15830.
Full text
31
Huang, Yumei. "Numerical methods for image restoration." HKBU Institutional Repository, 2008. http://repository.hkbu.edu.hk/etd_ra/908.
Full text
32
Yan, Ruomei. "Adaptive representations for image restoration." Thesis, University of Sheffield, 2014. http://etheses.whiterose.ac.uk/6975/.
Full textAbstract:
In the field of image processing, building good representation models for natural images is crucial for various applications, such as image restoration, sampling, segmentation, etc. Adaptive image representation models are designed for describing the intrinsic structures of natural images. In the classical Bayesian inference, this representation is often known as the prior of the intensity distribution of the input image. Early image priors have forms such as total variation norm, Markov Random Fields (MRF), and wavelets. Recently, image priors obtained from machine learning techniques tend to be more adaptive, which aims at capturing the natural image models via learning from larger databases. In this thesis, we study adaptive representations of natural images for image restoration. The purpose of image restoration is to remove the artifacts which degrade an image. The degradation comes in many forms such as image blurs, noises, and artifacts from the codec. Take image denoising for an example. There are several classic representation methods which can generate state-of-the-art results. The first one is the assumption of image self-similarity. However, this representation has the issue that sometimes the self-similarity assumption would fail because of high noise levels or unique image contents. The second one is the wavelet based nonlocal representation, which also has a problem in that the fixed basis function is not adaptive enough for any arbitrary type of input images. The third is the sparse coding using over-complete dictionaries, which does not have the hierarchical structure that is similar to the one in human visual system and is therefore prone to denoising artifacts. My research started from image denoising. Through the thorough review and evaluation of state-of-the-art denoising methods, it was found that the representation of images is substantially important for the denoising technique. At the same time, an improvement on one of the nonlocal denoising methods was proposed, which improves the representation of images by the integration of Gaussian blur, clustering and Rotationally Invariant Block Matching. Enlightened by the successful application of sparse coding in compressive sensing, we exploited the image self-similarity by using a sparse representation based on wavelet coefficients in a nonlocal and hierarchical way, which generates competitive results compared to the state-of-the-art denoising algorithms. Meanwhile, another adaptive local filter learned by Genetic Programming (GP) was proposed for efficient image denoising. In this work, we employed GP to find the optimal representations for local image patches through training on massive datasets, which yields competitive results compared to state-of-the-art local denoising filters. After successfully dealing with the denoising part, we moved to the parameter estimation for image degradation models. For instance, image blur identification uses deep learning, which has recently been proposed as a popular image representation approach. This work has also been extended to blur estimation based on the fact that the second step of the framework has been replaced with general regression neural network. In a word, in this thesis, spatial correlations, sparse coding, genetic programming, deep learning are explored as adaptive image representation models for both image restoration and parameter estimation. We conclude this thesis by considering methods based on machine learning to be the best adaptive representations for natural images. We have shown that they can generate better results than conventional representation models for the tasks of image denoising and deblurring.
33
Sandor, Viviana. "Wavelet-based digital image restoration." W&M ScholarWorks, 1998. https://scholarworks.wm.edu/etd/1539623937.
Full textAbstract:
Digital image restoration is a fundamental image processing problem with underlying physical motivations. A digital imaging system is unable to generate a continuum of ideal pointwise measurements of the input scene. Instead, the acquired digital image is an array of measured values. Generally, algorithms can be developed to remove a significant part of the error associated with these measure image values provided a proper model of the image acquisition system is used as the basis for the algorithm development. The continuous/discrete/continuous (C/D/C) model has proven to be a better alternative compared to the relatively incomplete image acquisition models commonly used in image restoration. Because it is more comprehensive, the C/D/C model offers a basis for developing significantly better restoration filters. The C/D/C model uses Fourier domain techniques to account for system blur at the image formation level, for the potentially important effects of aliasing, for additive noise and for blur at the image reconstruction level.;This dissertation develops a wavelet-based representation for the C/D/C model, including a theoretical treatment of convolution and sampling. This wavelet-based C/D/C model representation is used to formulate the image restoration problem as a generalized least squares problem. The use of wavelets discretizes the image acquisition kernel, and in this way the image restoration problem is also discrete. The generalized least squares problem is solved using the singular value decomposition. Because image restoration is only meaningful in the presence of noise, restoration solutions must deal with the issue of noise amplification. In this dissertation the treatment of noise is addressed with a restoration parameter related to the singular values of the discrete image acquisition kernel. The restoration procedure is assessed using simulated scenes and real scenes with various degrees of smoothness, in the presence of noise. All these scenes are restoration-challenging because they have a considerable amount of spatial detail at small scale. An empirical procedure that provides a good initial guess of the restoration parameter is devised.
34
Jammal, Ghada. "Multiscale image restoration in nuclear medicine." Phd thesis, [S.l.] : [s.n.], 2001. http://elib.tu-darmstadt.de/diss/000100/GJammal.pdf.
Full text
35
May, Kaaren Lonna. "Blind image restoration via constrained optimisation." Thesis, Imperial College London, 1999. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.313788.
Full text
36
Kwan, Chun-kit, and 關進傑. "Fast iterative methods for image restoration." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2000. http://hub.hku.hk/bib/B31224520.
Full text
37
Lee, Richard. "3D non-linear image restoration algorithms." Thesis, University of East Anglia, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.338227.
Full text
38
Morris, Robin David. "Image sequence restoration using Gobbs distributions." Thesis, University of Cambridge, 1995. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.387724.
Full text
39
Morris, Octavius John. "Image restoration using composite signal models." Thesis, Imperial College London, 1986. http://hdl.handle.net/10044/1/38111.
Full text
40
Pryce, Jonathan Michael. "The statistical mechanics of image restoration." Thesis, University of Edinburgh, 1993. http://hdl.handle.net/1842/12805.
Full textAbstract:
Image restoration is concerned with the recovery of an 'improved' image from a corrupted picture, utilizing a prior model of the source and noise processes. We present a Bayesian derivation of the posterior probability distribution, which describes the relative probabilities that a certain image was the original source, given the corrupted picture. The ensemble of such restored images is modelled as a Markov random field (Ising spin system). Using a prior on the density of edges in the source, we obtain the cost function of Geman and Geman via information theoretic arguments. Using a combination of Monte Carlo simulation, the mean field approximation, and series expansion methods, we investigate the performance of the restoration scheme as a function of the parameters we have identified in the posterior distribution. We find phase transitions separating regions in which the posterior distribution is data-like, from regions in which it is prior-like, and we can explain these sudden changes of behaviour in terms of the relative free energies of metastable states. We construct a measure of the quality of the posterior distribution and use this to explore the way in which the appropriateness of the choice of prior affects the performance of the restoration. The data-like and prior-like characteristics of the posterior distribution indicate the regions of parameter space where the restoration scheme is effective and ineffective respectively. We examine the question of how best to use the posterior distribution to prescribe a single 'optimal' restored image. We make a detailed comparison of two different estimators to determine which better characterizes the posterior distribution. We propose that the TPM estimate, based on the mean of the posterior, is a more sensible choice than the MAP estimate (the mode of the posterior), both in principle and in practice, and we provide several practical and theoretical arguments in support.
41
Kwan, Chun-kit. "Fast iterative methods for image restoration /." Hong Kong : University of Hong Kong, 2000. http://sunzi.lib.hku.hk:8888/cgi-bin/hkuto%5Ftoc%5Fpdf?B22956281.
Full text
42
Wu, Hsien-Huang. "Image restoration for improved spectral unmixing." Diss., The University of Arizona, 1992. http://hdl.handle.net/10150/186114.
Full textAbstract:
Because of the resolution limitations in remote sensing, the radiance recorded by the detector at each pixel is the integrated sum of the spectral radiance of all materials within the detector instantaneous-field-of-view (IFOV). If the detector IFOV covers more than one object class, the radiance detected is not characteristic of any single class but a mixture of all classes. These mixed pixels will have spectral signatures that fall within the convex hull formed by the signatures of all the classes. Traditional classifiers are therefore usually left with many misclassified or unclassified pixels. To remedy this problem, unmixing algorithms which decompose each pixel into a combination of several classes have been successfully applied to estimate the percentage of each class inside one pixel. In this dissertation, unmixing error of the least squares unmixing algorithm that is caused by the intrinsic data variance, system PSF blurring, detection noise, and band-to-band misregistration is analyzed and evaluated. For high unmixing accuracy, image restoration is proposed to remove the PSF blurring degradation. To objectively assess the restoration performance and expedite the design of our application-oriented restoration scheme, and objective criterion based on the measurement of spectral fidelity in frequency domain is suggested. Based on this criterion, a detailed comparison between the conventional Wiener filter and sampled Wiener filter is conducted, which highlights the significance of sampling aliasing and verifies the results obtained visually by other researchers. Our study shows that contrary to restoration for visual purposes, a partial restoration scheme, instead of full restoration, should be used for a better unmixing performance. Also, the sampling aliasing, which is an artifact and should be suppressed in traditional restoration application, is actually a signal component which needs to be restored for unmixing. Under fair SNR conditions ($\ge$30dB), the proposed restoration scheme can reduce the total unmixing error up to 40% to 70% depending on the scene complexity.
43
Miller, Casey Lee. "Image restoration using trellis-search methods." Diss., The University of Arizona, 1999. http://hdl.handle.net/10150/288963.
Full textAbstract:
Methods for the restoration of images corrupted by blur and noise are presented. During transmission through an optical or electrical channel, images become corrupted by blur and noise as a result of channel limitations (i.e. optical aberrations or a bandlimit). If images are treated as a matrix whose elements (pixels) assume a finite number of values then there is a large but finite set of possible images that can be transmitted. By treating this finite set as a 'signal' set, digital communications methods may be used to estimate the uncorrupted image given a blurred and noisy version. Specifically, row-by-row estimation, decision-feedback and vector-quantization are used to extend the 1D sequence estimation ability of the a-posteriori probability (APP) and Viterbi algorithm (VA) to the estimation of 2D images. Simulations show the 2D VA and APP algorithms return near-optimal estimates of binary images as well as improved estimates of greyscale images when compared with the conventional Wiener filter (WF) estimates. Unlike the WF, the VA and APP algorithms are shown to be capable of super-resolution and adaptable for use with signal-dependent Poisson noise corruption. Restorations of experimental data gathered from an optical imaging system are presented to support simulation results.
44
Hazra, Rajeeb. "Constrained least-squares digital image restoration." W&M ScholarWorks, 1995. https://scholarworks.wm.edu/etd/1539623865.
Full textAbstract:
The design of a digital image restoration filter must address four concerns: the completeness of the underlying imaging system model, the validity of the restoration metric used to derive the filter, the computational efficiency of the algorithm for computing the filter values and the ability to apply the filter in the spatial domain. Consistent with these four concerns, this dissertation presents a constrained least-squares (CLS) restoration filter for digital image restoration. The CLS restoration filter is based on a comprehensive, continuous-input/discrete- processing/continuous-output (c/d/c) imaging system model that accounts for acquisition blur, spatial sampling, additive noise and imperfect image reconstruction. The c/d/c model-based CLS restoration filter can be applied rigorously and is easier to compute than the corresponding c/d/c model-based Wiener restoration filter. The CLS restoration filter can be efficiently implemented in the spatial domain as a small convolution kernel. Simulated restorations are used to illustrate the CLS filter's performance for a range of imaging conditions. Restoration studies based, in part, on an actual Forward Looking Infrared (FLIR) imaging system, show that the CLS restoration filter can be used for effective range reduction. The CLS restoration filter is also successfully tested on blurred and noisy radiometric images of the earth's outgoing radiation field from a satellite-borne scanning radiometer used by the National Aeronautics and Space Administration (NASA) for atmospheric research.
45
Chana, Deeph S. "Image restoration exploiting statistical models of the image capture process." Thesis, King's College London (University of London), 2001. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.246886.
Full text
46
Veldhuizen, Todd Lawrence. "Grid filters for local nonlinear image restoration /." Waterloo, Ont. : University of Waterloo [Dept. of Systems Design Engineering], 1998. http://etd.uwaterloo.ca/etd/tveldhui1998.pdf.
Full textAbstract:
Thesis (M.A.Sc.)-University of Waterloo, 1998.Includes bibliographical references (leaves 109-115). Issued also in PDF format and available via the World Wide Web. Requires Internet connectivity, World Wide Web browser, and Adobe Acrobat Reader.
47
Veldhuizen, Todd. "Grid Filters for Local Nonlinear Image Restoration." Thesis, University of Waterloo, 1998. http://hdl.handle.net/10012/943.
Full textAbstract:
A new approach to local nonlinear image restoration is described, based on approximating functions using a regular grid of points in a many-dimensional space. Symmetry reductions and compression of the sparse grid make it feasible to work with twelve-dimensional grids as large as 2212. Unlike polynomials and neural networks whose filtering complexity per pixel is linear in the number of filter co-efficients, grid filters have O(1) complexity per pixel. Grid filters require only a single presentation of the training samples, are numerically stable, leave unusual image features unchanged, and are a superset of order statistic filters. Results are presented for additive noise, blurring, and superresolution.
48
Langari, Bahareh. "Multi-scale edge-guided image gap restoration." Thesis, Brunel University, 2016. http://bura.brunel.ac.uk/handle/2438/13406.
Full textAbstract:
The focus of this research work is the estimation of gaps (missing blocks) in digital images. To progress the research two main issues were identified as (1) the appropriate domains for image gap restoration and (2) the methodologies for gap interpolation. Multi-scale transforms provide an appropriate framework for gap restoration. The main advantages are transformations into a set of frequency and scales and the ability to progressively reduce the size of the gap to one sample wide at the transform apex. Two types of multi-scale transform were considered for comparative evaluation; 2-dimensional (2D) discrete cosines (DCT) pyramid and 2D discrete wavelets (DWT). For image gap estimation, a family of conventional weighted interpolators and directional edge-guided interpolators are developed and evaluated. Two types of edges were considered; ‘local’ edges or textures and ‘global’ edges such as the boundaries between objects or within/across patterns in the image. For local edge, or texture, modelling a number of methods were explored which aim to reconstruct a set of gradients across the restored gap as those computed from the known neighbourhood. These differential gradients are estimated along the geometrical vertical, horizontal and cross directions for each pixel of the gap. The edge-guided interpolators aim to operate on distinct regions confined within edge lines. For global edge-guided interpolation, two main methods explored are Sobel and Canny detectors. The latter provides improved edge detection. The combination and integration of different multi-scale domains, local edge interpolators, global edge-guided interpolators and iterative estimation of edges provided a variety of configurations that were comparatively explored and evaluated. For evaluation a set of images commonly used in the literature work were employed together with simulated regular and random image gaps at a variety of loss rate. The performance measures used are the peak signal to noise ratio (PSNR) and structure similarity index (SSIM). The results obtained are better than the state of the art reported in the literature.
49
Palmer, Alexander S. "Adaptive image restoration algorithms using intelligent techniques." Thesis, University of East Anglia, 2003. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.405233.
Full text
50
Talebi-Rafsanjan, Siamak. "Image restoration techniques for bursty erasure channels." Thesis, King's College London (University of London), 2003. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.406409.
Full text